Medical device surveillance with electronic health records
Authors:
Alison Callahan,
Jason A Fries,
Christopher Ré,
James I Huddleston III,
Nicholas J Giori,
Scott Delp,
Nigam H Shah
Abstract:
Post-market medical device surveillance is a challenge facing manufacturers, regulatory agencies, and health care providers. Electronic health records are valuable sources of real world evidence to assess device safety and track device-related patient outcomes over time. However, distilling this evidence remains challenging, as information is fractured across clinical notes and structured records.…
▽ More
Post-market medical device surveillance is a challenge facing manufacturers, regulatory agencies, and health care providers. Electronic health records are valuable sources of real world evidence to assess device safety and track device-related patient outcomes over time. However, distilling this evidence remains challenging, as information is fractured across clinical notes and structured records. Modern machine learning methods for machine reading promise to unlock increasingly complex information from text, but face barriers due to their reliance on large and expensive hand-labeled training sets. To address these challenges, we developed and validated state-of-the-art deep learning methods that identify patient outcomes from clinical notes without requiring hand-labeled training data. Using hip replacements as a test case, our methods accurately extracted implant details and reports of complications and pain from electronic health records with up to 96.3% precision, 98.5% recall, and 97.4% F1, improved classification performance by 12.7- 53.0% over rule-based methods, and detected over 6 times as many complication events compared to using structured data alone. Using these events to assess complication-free survivorship of different implant systems, we found significant variation between implants, including for risk of revision surgery, which could not be detected using coded data alone. Patients with revision surgeries had more hip pain mentions in the post-hip replacement, pre-revision period compared to patients with no evidence of revision surgery (mean hip pain mentions 4.97 vs. 3.23; t = 5.14; p < 0.001). Some implant models were associated with higher or lower rates of hip pain mentions. Our methods complement existing surveillance mechanisms by requiring orders of magnitude less hand-labeled training data, offering a scalable solution for national medical device surveillance.
△ Less
Submitted 3 April, 2019;
originally announced April 2019.
